Diabetes is an increasing worldwide problem, with approximately a fourfold increase in prevalence over a thirty-year period in the latest World Health Organisation figures (1).
Dr David Ewins is a Consultant Physician in Endocrinology and Diabetes and talks with Scott about the approach to evaluating and managing adult patients presenting with high blood glucose levels in hospital, including the use of a Variable Rate Intravenous Insulin Infusion (VRIII).
In the UK, there are almost 3.6 million people currently diagnosed with diabetes and an estimated 1 million people with undiagnosed type 2 diabetes (2)with a trend of a continuing increase in prevalence. The 2016 national audit from England and Wales suggests diabetes affects 20% of hospital inpatients and that it can be poorly managed in hospital (3). This is known to result in adverse outcomes for patients including increased morbidity and mortality and longer hospital stays (4,5).
The main focus of the discussion in this podcast is the initial assessment and the treatment of high blood glucose levels on the wards including key points from the history, examination and important investigations.
We start by discussing a typical example case.
• 46 year old patient
• Background of Insulin treated diabetes
• Hospital inpatient with high blood glucose readings including 27 mmol/L intermittently over the last 3 days
• Previously erratic glucose control
A key question to assess is to ascertain whether the patient has type 1 diabetes, insulin treated type 2 diabetes, or type 2 diabetes treated with diet and/or oral medication (4).
This is important to establish early on because patients with type 1 diabetes are dependent upon insulin treatment which must generally be continued to prevent Diabetes Ketoacidosis (DKA). Whereas patients with insulin treated type 2 diabetes could potentially have their insulin treatment temporarily suspended whilst being assessed further. In addition to asking the patient, there may be clues as to the type of diabetes a patient has.
This includes a history of being treated with diet or tablets initially and later on insulin suggesting insulin treated type 2 diabetes. A family history or personal history in the patient of autoimmune disease including hashimoto’s thyroiditis, coeliac disease, or pernicious anaemia would be more likely to be consistent with type 1 diabetes (6,7).
Patients with an elevated Body Mass Index (BMI) who are overweight or obese would be more likely to have insulin treated type 2 diabetes. However, it is important to remember that all these clues are not diagnostic, and for example, there are overweight patients with type 1 diabetes so avoid being caught out. The patient’s medical records and previous letters or local records from their General Practitioner (GP) are often an invaluable tool for providing this information, and it is possible to see what medications a patient has been taking previously.
After establishing what type of diabetes the patient has, the next important thing to do is to assess the clinical state of the patient (4,6). This includes:
• General appearance- ‘well or unwell?’
• Are they eating and drinking?
• The Modified Early Warning (MEWS) score
• Are any features of Diabetic Ketoacidosis (DKA) present ? (including clinical or biochemical findings which are described in our seperate DKA blog and podcast)
• Is there a reason for their raised glucose level, including incurrent illness such as a pneumonia or urinary tract infection?
• Are there other biochemical derangements including the potential for Acute Kidney Injury (AKI)?
• Is there a contributing or alternate cause of the patient’s raised blood glucose besides diabetes including steroid treatment or Total Parenteral Nutrition (TPN)?
• What has the patient’s blood glucose trend been recently, including any hypoglycaemic episodes?
• Has the patient had surgery recently?
Working through the above questions can help provide the information to explain recent changes in a patient’s glucose level and help determine the correct management in line with the type of diabetes.
Managing hyperglycaemia in patients with type 2 diabetes
In patients with insulin treated type 2 diabetes presenting with a raised blood glucose, who are well and eating and drinking, it is important to check if they have been compliant with their usual insulin regime. If the patient has been compliant, with no recent doses of insulin omitted, then their usual insulin regime doses can be titrated upwards from their next dose to allow for better control of their diabetes (4,6,7).
Ideally, stat doses of actrapid or short acting insulin should be avoided in this situation. Firstly, due to the risk of causing subsequent hypoglycaemia. Secondly, the underlying insulin regime needs adjusting otherwise hyperglycaemia will tend to recur.
Unwell Type 2 Diabetic Patient with Hyperglycaemia
In patients with insulin treated type 2 diabetes who are unwell, including those who are not eating and drinking, it is important to accurately establish/confirm what their blood glucose is by checking their venous blood glucose, Urea and Electrolytes (U and Es), venous blood gas, ketones and serum osmolality. The serum osmolality can be measured directly by the laboratory in their venous blood or can be approximated from the formula for serum osmolarity (8,9):
Serum osmolarity = 2 (Sodium + Potassium) + urea + glucose
If the bicarbonate on the U and Es is low (<15mmol/L) and the ketones are raised (>3.0mmol/L) with a low pH (venous blood gas pH <7.3) this would suggest the presence of DKA. DKA can occur in type 1 diabetes, and is not common in type 2 diabetes but can occur in the context of illness. If present, DKA should be managed as a medical emergency according to your local DKA pathway. The management of DKA is discussed in the Joint British Diabetes Societies Guidelines (10,11) and here.
If the patient with insulin treated type 2 diabetes presents with very elevated blood glucose levels (typically > 30mmol/L or more), does not fit the criteria for DKA but their serum osmolality is high (>320mosmol/L) with evidence of hypovolaemia/dehydration, they should be treated for the medical emergency of Hyperosmolar Hyperglycaemic Syndrome (HHS) (4,12,13).
All suspected cases of DKA or HHS should be discussed with a senior doctor oncall and the diabetes team informed immediately for the provision of specialist advice.
HHS will require IV fluid resuscitation (13). The first choice initial fluid to use for this is sodium chloride 0.9%. The administration of IV fluids themselves will lead to a fall in blood glucose level, which is usually very elevated to begin with. Here, we are aiming for a fall in the blood glucose of between 4 to 6mmol/L/hour (ideally 5mmol/L/hour).
There is a risk in patients with HHS of commencing intravenous insulin prior to adequate fluid replacement, thereby causing a rapid lowering of their glucose level with a rapid fall in serum osmolality and cardiovascular collapse as the insulin drives glucose into their cells (with water moving out of the intravascular compartment with glucose into cells), and glucose is excreted (with water) through the kidneys in an osmotic diuresis. Most patients with HHS are insulin sensitive.
Therefore, ketones should be checked (point of care hospital testing is now available- e.g. in A&E/MAU/check to borrow a meter from a diabetes ward and bring it back) and if > 1mmol/L (this would indicate a relative hypoinsulinaemia), then IV insulin should be commenced in a Fixed Rate Intravenous Insulin Infusion with IV fluids at time zero (similarly to DKA, although generally lower doses are required). If IV insulin is commenced in an FRIII, the JBDS recommend using 0.05units/kg/hour (which equates to 4 units per hour for an 80Kg patient for example).
If the ketones are < 1mmol/L, IV insulin is not required in the immediate phase, and the patient should be adequately fluid resuscitated first.
If the blood gluocose is not falling by 5mmol/L/hr with IV fluid resucitation (in a patient with initial serum ketones < 1mmol/L), their fluid balance and renal function should be re assessed and IV insulin in an FRIII can be commenced at this point (or if an FRIII is already running- in a patient with initial serum ketones > 1mmol/L, the rate of the FRIII here can be increased by 1unit/hour).
Again we are aiming for a fall in blood glucose of ideally 5mmol/L/hour, with a target blood glucose with treatment in this instance of 10 to 15mmol/L. It can take up to 72 hour for their blood glucose and serum electrolytes to fully normalise.
As with DKA, in HHS potassium will need to be checked regularly on the patient’s Urea and Electrolytes (with the patient’s other blood tests for Glucose, Ketones, Venous Blood Gas- at time zero, 1hr, 4hr, 6hr, 12hr) with potassium being replaced in IV fluids, although because patients with HHS are less acidotic (in DKA correction of the acidosis drives potassium into their cells) than those with DKA, and the insulin dose used is lower (insulin drives potassium into cells also), in HHS potassium shifts tend to be less rapid.
In HHS, patients could present with an Acute Kidney Injury with Hyperkalaemia, or could be profoundly hypokalaemic if on diuretics (with potassium loss) for example. Therefore, the JBDS guidelines for HHS recommend checking a potassium level (regular in the initial phase of treatment as for DKA) and replacing or omitting it in their IV fluids (0.9% normal saline +/- potassium) as required- this suggests for a potassium level of 5.5mmol/L- no potassium replacement in the fluid bag, for potassium level 3.5 to 5.5mmol/L to add 40mmol/L to the fluid bag, and for a potassium level below 3.5 for senior review as additional potassium may be required via a central line in HDU. This replacement will be guided by your local hospital policy.
In addition, prophylactic dose Low Molecular Weight Heparin (LMWH) e.g. tinzaparin is provided for the full duration of the admission. This is to guard against the risk of venous thromboembolism in the hyperosmolar dehydrated state of HHS. This can lead to a hypercoagulable state, increasing the risk of clots occurring in the circulation.
Extending the length of administration of prophylactic LMWH beyond the length of admission (for example by 1 week) in patients felt to be at particularly high risk of VTE should be considered in those treated for HHS.
There is some debate regarding the evidence for the use of therapeutic or prophylactic LMWH for HHS and this will be guided by your local hospital policy (13,14,15). Prophylactic dose LMWH is often given, but particularly if there is evidence of co existent Acute Coronary Syndrome (ACS) (which may have precipitated the HHS) or thrombosis then treatment dose LMWH would be given- check your local hospital policy.
As for DKA, any features of infection should be sought with an infection screen (including CXR, urine dipstick -MSU, checking for any cellulitis or other causes) and promptly treated with antibiotics as appropriate- as the infection may be driving the HHS. It should be remembered that septic patients are not always pyrexial!
If a patient with insulin treated type 2 diabetes neither has DKA nor HHS, but their blood glucose level is high and they are unwell with limited or no oral intake, a VRIII should be commenced as this is the safest option to bring their glucose under control as the primary cause of the raised blood glucose is treated e.g. intercurrent illness, infection, myocardial infarction (16,17).
Managing hyperglycaemia in patients with type 1 diabetes
If a patient with type 1 diabetes presents with raised blood glucose levels and is well, with normal oral intake and no features of DKA, then their usual insulin regime can be titrated upwards from the next scheduled dose to allow for better control of their diabetes (4).
With mild hyperglycaemia, stat doses of actrapid or short acting insulin should ideally be avoided because of the risk of subsequent hypoglycaemia and recurring hyperglycaemia if the underlying insulin regime is not adjusted. However, if the blood glucose level is high > 20mmol/L, giving a correction dose of short acting insulin with the patient’s next meal can be considered, and it is helpful to ask the patient if they have used correction doses in the past and what dose they have taken.
As a guide if the patient is unsure or has not used a correction dose before, a dose of 1 unit of insulin will bring the blood glucose down by 3mmol/L so the correction dose can be calculated aiming to bring the blood glucose back within the target acceptable blood glucose range (4 to 12 mmol/L).
For example, with a blood glucose of 21mmol/L, a correction dose of 4 units of short acting insulin could be considered with the patient’s next meal. This would be expected to bring the patient’s glucose down by approximately 12mmol/L (4 times 3) to 9mmol/L which is then within the target range (4,9).
If a patient with type 1 diabetes presents with raised blood glucose levels and is unwell, with limited or no oral intake, then as for patients in this situation with type 2 diabetes, a venous blood glucose, venous blood gas, U and Es, and blood ketones sample should be taken.
If their blood ketones are not elevated (<1.5mmol/L), a correction dose of insulin with their next meal can be considered. The patient’s capillary blood glucose readings should be monitored closely, half an hour to an hour after giving the correction dose to ensure that they are not becoming hypoglycaemic (18,19).
If the ketones are raised (>3.0mmol/L), and the bicarbonate on the U and Es is low (<15mmol/L) with a low pH (venous blood gas pH <7.3) this would suggest the presence of DKA. Again, if present, DKA should be managed according to your local DKA pathway. The management of DKA is discussed in the Joint British Diabetes Society Guidelines (10,11) and here.
If the ketones are slightly elevated above 1mmol/L, but the patient is not acidotic and therefore does not meet the criteria for DKA, a VRIII should be considered.
Variable Rate Insulin Infusion
A Variable Rate Intravenous Insulin Infusion (VRIII) is an infusion of insulin, made up in a syringe typically with 50 units of short acting insulin e.g. actrapid in 50ml of normal 0.9% sodium chloride solution (17). This is the same infusion syringe as that used in DKA. Unlike in DKA where this infusion is a fixed rate intravenous insulin infusion (FRIII), for a VRIII the rate of insulin infused from this syringe is altered on the syringe driver pump according to hourly capillary blood glucose measurements. When using an FRIII in DKA, separate intravenous fluids are given according to the DKA protocol (10).
When using a VRIII, in parallel to the VRIII usually through a Y connector, an intravenous fluid substrate should be run which is usually 5% dextrose with 0.45% sodium chloride with 10 to 20mmol of potassium chloride as recommended by the Joint British Diabetes Society (17). This will be guided by your local hospital protocol for the prescription of a VRIII. This is typically run at a 100ml/hr with the rate adjusted according to the fluid balance requirements of the individual patient and clinical judgement.
A VRIII prescription chart may have a standard table column on the chart for altering the insulin infusion rate depending upon the hourly capillary blood glucose reading. In patients who are known to be insulin sensitive and prone to episodes of hypoglycaemia, often in patients with a total daily insulin dose of less than 24 units of insulin per day, a reduced insulin infusion rate on the VRIIII may be used. Depending on your local policy and the individual patient, this may be shown on the chart as an ‘insulin sensitive or decreased’ column to be used in these cases (17).
By comparison, in overweight patients with insulin resistance and typically a daily insulin dose of more than 100 units of insulin per day, these patients may require higher doses of insulin. Depending on your local policy this can be seen on some VRIII charts as an ‘insulin resistant or increased’ column to be used in these patients.
Occasionally the VRIII adjustment table can be manually altered by the inpatient diabetes team to suit the insulin requirements of an individual patient.
VRIII versus “Sliding Scale Insulin”
Insulin sliding scale is an old term which is an umbrella term for any insulin dose which is adjusted according to a blood glucose reading with a table. The term does not delineate between whether the insulin is being given subcutaneously or intravenously and does not describe the composition of any fluids given concurrently. It is therefore vague and means different things to different people and the term should be avoided. It is safer to refer to the variable rate of intravenous insulin as a VRIII to avoid any confusion.
It is possible to be over reactive to raised blood glucose readings in patients by prescribing several correction doses of insulin. This should be avoided as it runs the risk of hypoglycaemia. It also does not correct the underlying basal insulin dose which will mean the raised blood glucose is likely to keep occurring. Correction doses should not be given more often than every 4 hours and only if required (4).
Assess the patient for any clinical signs or symptoms of hypoglycaemia (capillary glucose < 4mmol/L) including feeling hungry, tired or agitated/aggressive and appearing clammy, sweating, shaking or evidence of confusion (6,7). As the blood glucose falls further in severe cases, or if the initial features have not been spotted and treated, a reducing level of consciousness into coma can occur as well as seizures and there is a risk of brain damage (6,7,11).
If the patient is alert, awake and able to swallow safely, 15 to 20g of fast acting carbohydrate should be given orally. This may include a glass of milk with two spoons of sugar, sweets, glucose gel, a sugary drink or pure fruit juice. If the patient is drowsy, glucose gel may be given by the nursing staff. If the patient is unable to swallow or there are any concerns, intravenous 10% dextrose should be given e.g. 250ml (equivalent to 25g of glucose) stat.
This may be repeated if required. An intravenous maintenance dextrose 10% infusion may be required for prolonged hypoglycaemia. This is titrated to maintain their capillary blood glucose in the target range of 4 to 12mmol/L. When the patient is conscious and safely able to swallow, sugary drinks and a meal should be provided (9,11,12).
Continuing Long Acting Insulin
If patients are already taking a long acting subcutaneous insulin analogue (e.g. Lantus or Levemir), this should continue to be prescribed and administered once daily whilst they are on a VRIII as this will make switching the patient back to their usual subcutaneous insulin regime much easier when they are well enough to come off the VRIII by providing some background insulin (17).
Patients receiving insulin pump therapy can equally become unwell in the same way as patients without an insulin pump. These patients should therefore be managed in the same general approach.
If a patient with an insulin pump has a raised blood glucose reading but is otherwise well and managing to eat and drink, a correction dose of insulin given by the insulin pump may be appropriate. The patient and inpatient specialist diabetes team have knowledge of how to operate their pump and should be contacted.
If a patient with an insulin pump has a raised blood glucose reading and is unwell with limited or no oral intake, which would require the use of a VRIII or treatment as DKA with an FRIII, then their insulin pump should be stopped, disconnected and they should be treated with a VRIII (or FRIII for DKA) as appropriate.
Only the patient and inpatient specialist diabetes team would be able to operate their insulin pump and therefore would need to be contacted in any case.
Transferring from VRIII to Regular Insulin
When the patient well, biochemically back to normal and is eating and drinking, provided that their long acting subcutaneous insulin has been continued, their short acting insulin can then be restarted at their usual dose just before their next meal (17).
The VRIII can be stopped approximately 30 minutes after giving this. This overlap period is the safest way of transitioning by reducing the risk of a rebound high blood glucose upon stopping the VRIII.
The patient will then continue their usual insulin regime e.g. basal bolus, with short acting subcutaneous insulin given with subsequent meals and their long acting subcutaneous insulin continued throughout.
If possible, it is best to attempt to switch the patient back to their usual subcutaneous insulin during the day (e.g. before breakfast or lunch, rather than before their evening meal). This helps ensure continuity of care should there be subsequent fluctuations in their blood glucose level.
Referral to Diabetes Specialist Teams
Any inpatient with diabetes who is unwell should be referred to the inpatient specialist diabetes team, including the diabetes specialist nurse at the earliest opportunity.
In summary, we have discussed the assessment and management of patients with Diabetes Mellitus with high blood glucose levels including the use of a VRIII. A reference which is essential further reading is ‘Managing adults with diabetes in hospital during an acute illness’ from the BMJ, June 2017 (4).
I’d like to thanks Dr Ewins, Consultant Physician in Endocrinology and Diabetes, for his time and effort in recording this podcast, and the medical education department at Chester for their continued support.
GET IN TOUCH! Either through the comments section or email firstname.lastname@example.org if there are any comments or corrections or if you want to be involved with future topics.
RATE US on iTunes to help others find the resource more easily!
SHARE US with others you think would benefit or even your training leads to help us distribute the resource further! Twitter: @FOAMdation @scottymwilliams
- World Health Organisation Media Centre Diabetes Fact Sheet. http://www.who.int/mediacentre/factsheets/fs312/en/ Updated July 2017. Accessed 24/7/17
- Diabetes UK position statement diabetes prevalence 2016. https://www.diabetes.org.uk/Professionals/Position-statements-reports/Statistics/Diabetes-prevalence-2016/ Accessed 21/7/17
- NADIA Team. National Diabetes In-patient Audit, England and Wales, 2016. NHS Digital, March 2016
- Chowdhury et al. Managing adults with diabetes in hospital during an acute illness. BMJ 2017;357:528-531
- Evans, Dhatariya. Assessing the relationship between admission glucose levels, subsequent length of hospital stay, readmission and mortality. Clin Med (Lond) 2012;357:137-9
- Kasper, Fauci et al. Harrison’s Principles of Internal Medicine 19th edition. McGrawhill 2016
- Kumar, Clark et al. Kumar and Clark’s Clinical Medicine 9th edition. Elsevier 2016
- Up To Date Online, Serum Osmolality Formula. https://www.uptodate.com/contents/serum-osmolal-ga Accessed 3/7/17
- Longmore, Wilkinson et al. Oxford Handbook of Clinical Medicine 9th edition. Oxford University Press 2014
- The Management of Diabetic Ketoacidosis in Adults. Joint British Diabetes Societies Inpatient Care Group. Second Edition Update: September 2013. Accessed 5/7/17
- NICE Guideline NG17 Type 1 Diabetes in adults: diagnosis and management. Updated July 2016. Accessed 6/7/17
- NICE Clinical Guideline NG 28 Type 2 Diabetes in adults: management. Updated May 2017, Accessed 23/7/17
- The management of the hyperosmolar hyperglycaemic state (HHS) in adults with diabetes. Joint British Diabetes Societies Inpatient Care Group 2012.
- Silvio and Inzucchi. Management of hyperglycemia in the hospital setting. New England Journal of Medicine 2006;357:1903-11
- Wordsworth, Robinson et al. HHS – full or prophylactic anticoagulation? Br J Diabetes Vasc Dis 2014;14:64-66
- Kitabchi, Umpierrez et al. Hyperglycemic crises in adult patients with diabetes. Diabetes Care 2009;357:1335-43
- Joint British Diabetes Societies for inpatient care. The use of variable rate intravenous insulin infusion (VRIII) for medical inpatients 2014. https://www.diabetes.org.uk/Documents/About%20Us/What%20we%20say/Use%20of%20variable%20rate%20intravenous%20insulin%20infusion%20in%20medical%20inpatients.pdf Accessed 22/7/17
- NICE Clinical Guideline CG 50 Acutely ill patients in hospital. Updated 2013,
- Marik. Tight glycemic control in acutely ill patients: low evidence of benefit, high evidence of harm! Intensive Care Medicine 2016;357:1475-7